publisher's layout or pagination. thomas, e.f. , mcgarty ... · lala, smith, & bliuc,...

MURDOCH RESEARCH REPOSITORY

http://researchrepository.murdoch.edu.au

This is the author's final version of the work, as accepted for publication following peer review but without thepublisher's layout or pagination.

Thomas, E.F. , McGarty, C. , Lala, G. , Stuart, A. , Hall, L.J. and Goddard, A. (2015) Whatever happened toKony2012? Understanding a global Internet phenomenon as an emergent social identity. European

Journal of Social Psychology, 45 (3). pp. 256-367.

http://researchrepository.murdoch.edu.au/25984

Copyright © Â© 2015 John Wiley & Sons, Ltd.It is posted here for your personal use. No further distribution is permitted.

http://researchrepository.murdoch.edu.au

http://researchrepository.murdoch.edu.au/25984

UNDERSTANDING KONY 2012

Whatever Happened to Kony2012? Understanding a Global Internet Phenomenon as an

Emergent Social Identity

Emma F. Thomas1

Craig McGarty1, 2

Girish Lala1

Avelie Stuart1, 3

Lauren J. Hall1

Alice Goddard1

1 School of Psychology and Exercise Science, Murdoch University

2 School of Social Sciences and Psychology, University of Western Sydney

3 School of Psychology, University of Exeter

Address for correspondence: Emma F Thomas, School of Psychology & Exercise

Science, Murdoch University, Murdoch, 6150, WA, Australia. E-mail:

[email protected]. The authors would like to acknowledge Yara Neto for her

assistance with data collection. This research was supported by the Australian Research

Council’s Discovery scheme (DP110100036) and Discovery Early Career Researcher Award

(DE120101029).

UNDERSTANDING KONY2012

2

Abstract

Kony2012 was a viral Internet video that attracted unprecedented online interest in promoting

a campaign to arrest the leader of an African militant group. The current research considers

the social psychological bases of social media-based collective action. In three cross sectional

surveys (N = 304) collected before, on, and after the key action date of April 20, 2012, we

consider the nature (opinion-based or global) and function (emergent or transforming) of

social identity in modern forms of social action. Multigroup structural equation modeling

showed that Kony2012 action was best captured by an emergent opinion-based social

identity. Moreover, the same factors that predicted Kony2012 action generally also predicted

engagement in new repertoires of protest (involving the use of social media) and an

observable traditional socio-political action (signing a letter to a government minister). The

results suggest that there is no sharp dividing line between traditional and new forms of

collective action and that both may be understood as valid expressions of collective selfhood.


3

Whatever Happened to Kony2012? Understanding a Global Internet Phenomenon as an

Emergent Social Identity.

Perhaps digital media may foster a new kind of collective identity—a virtual one—that is in some ways

more effectively mobilizing than the kind of collective identity forged in church basements and college

dorm lounges by people holding hands and singing We Shall Overcome (Polletta, 2011).

On March 5, 2012 the US-based advocacy group Invisible Children released a 30

minute YouTube video calling for action against the Lord’s Resistance Army (LRA). The

organization’s aim was to make LRA leader Joseph Kony accountable by making him

(in)famous and thus to raise international support for his arrest (www.Kony2012.com).

Supporters of Kony2012 were asked to purchase campaign kits, contact decision makers and

to cover walls throughout the world with posters on April 20, 2012 (termed ‘Cover the

Night’), in an effort to draw political attention to Kony’s continued activities. Supporters were

also urged to share the YouTube video. In that aspect the film was spectacularly successful,

attracting approximately 100 million viewings. Time magazine ranked the film as the most

viral video of all time (Time, 2012), and one poll suggested that more than 50% of young

American adults had heard about Kony2012 in the days following the release of the film (Pew

Research Center, 2012). However, not long after the film’s popularity reached its zenith, the

campaign’s legitimacy was questioned (specifically, questions were raised about the

appropriate use of funds) and the film’s director attracted negative publicity through a public

breakdown (see Curtis & McCarthy, 2012). Cover the Night attracted a low level of support

(e. g., News, 2012) and critics of ‘slacktivism’ (or ‘clicktivism’) pointed to Kony2012 as

another example of the failures of online activism (e.g. Bailyn, 2012, Mengestu, 2012; also

Gladwell, 2010, for a discussion of the broader point).


4

The juxtaposition between Kony2012 and the Arab Spring protests is stark. The low

levels of offline mobilization arising from Kony2012 stand in contrast to the rapid

progression of the Arab Spring of late 2010 and 2011 where social media played a vital role

in coordinating offline action (Howard et al., 2011; Lotan et al., 2011; McGarty, Thomas,

Lala, Smith, & Bliuc, 2014). Indeed, McGarty et al. (2014) showed that online interaction set

the scene for the massive displays of offline action that came to characterize those

revolutions. In contrast, the much longer Kony2012 video was seen by many more people but

generated very little action on the street. Are online mobilization and traditional socio-

political action qualitatively different phenomena that need to be explained in different ways,

or are they aspects of the same thing? One way of addressing this complex question is to ask

whether those two forms of action share the same social psychological underpinnings.

In this paper we draw on the frameworks provided by two recent models of collective

action to examine Kony2012: the social identity model of collective action (SIMCA; van

Zomeren, Postmes, & Spears, 2008), and the encapsulated model of social identity in

collective action (EMSICA; Thomas, Mavor, & McGarty, 2012; Thomas, McGarty, &

Mavor, 2009). Both models articulate a prominent role for social identity, reactions to

injustice, and group efficacy beliefs (that is, the belief that coordinated action can be

effective; Bandura, 2000), in motivating action for social change. A key question here is

whether these frameworks adequately explain emergent global internet phenomena such as

Kony2012, movements that are overwhelmingly characterized by what has been called ‘the

new repertoire of protest’ (Polletta, 2011). Specifically, while some have questioned whether

social identities are even necessary to understand modern (primarily online) social

movements (e.g. Bennett & Segerberg, 2012; Earl & Kimport, 2011), we follow Polletta

(2011, above) and seek to advance an analysis of the nature and function of social identity in

these global online social movements.


5

Social Identity, Injustice, and Efficacy in Kony2012

Social identities are understood to be central to coordinated, group-level social action

(Reicher, 1984; see van Zomeren et al., 2009 for a meta-analysis; Thomas et al., 2009, for a

review) but a key issue raised by movements such as Kony2012 is that there were no existing

politicized collective identities (Simon & Klandermans, 2001) or manifest ingroup

disadvantages amongst the citizens of wealthy nations (van Zomeren, Leach, & Spears, 2012).

Unlike many other social movements such as civil rights movements or the gay and lesbian

rights movements, the participants do not themselves experience the structural disadvantage

or repression (they are primarily relatively advantaged; see Leach, Snider & Iyer, 2002). How

does one promote united collective action on behalf of people from other countries, where

there is no common threat, incentive or pre-existing social identity (a problem considered by,

for example, Nadler & Halabi, 2006; Reicher, Hopkins, Levine, & Rath, 2005; Thomas et al.,

2009)? Here the literature suggests two alternatives.

The Kony2012 campaign entreated citizens around the world to help African children

and thus it could be seen as an attempt to appeal to a superordinate global or human social

identity. According to the tenets of self-categorization theory, people who identify with their

global community should be motivated to support the campaign against Kony because doing

so would advance the cause of humanity generally and fellow human beings in particular

(Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). In a multinational study Buchan et al.

(2011) showed that global identification was an independent predictor of global cooperation

and global concerns. Specifically, Buchan et al. demonstrated that identification with the

people of the world predicted behavioural contributions to a public goods dilemma,

transcending parochial interests. In a separate line of research, McFarland, Webb and Brown

(2012) highlight the role of identification with all of humanity in predicting knowledge of


6

(Study 8) and learning about (Study 9) global humanitarian concerns, and a willingness to

contribute to international humanitarian relief (Study 10). Thus, research in the context of

global cooperation (but not collective action specifically; Buchan et al., 2011) and global

human rights (McFarland et al., 2012; McFarland, Brown & Webb, 2013) suggests that a

global, superordinate identity may underpin Kony2012 action. Indeed, in the case of

Kony2012, the imagery used in the campaign drew upon global identity with depictions of the

problem as something to be addressed by united action with other people from all over the

world (see Invisible Children, 2012; also Waldorf, 2013).

However, other literature suggests that the relationship between global identification

and mobilization in online movements such as Kony2012 may not be straightforward. For

example, Greenaway, Quinn and Louis (2011) showed that appeals to common humanity

promoted forgiveness but lowered collective action intentions amongst victims of historical

atrocities: If we are all fellow human beings why would we need to act against one another?

Similarly, Morton and Postmes (2011) suggest that invoking a common humanity can serve

as a moral defence for harmful intergroup activity; we are all ‘only human’. Finally, Reese,

Berthold and Steffens (2012) showed that members of wealthy countries project their own

group attributes onto superordinate global categories, such that ‘identification with the world’

is much more akin to ‘identification with advantaged people like me’. It is also the case that

such global or superordinate identities are unlikely to be easily politicized (Simon &

Klandermans, 2001). As politicized identities – those embedded in a context which

recognizes a system of unequal power relations – are stronger predictors of collective action

than are non-politicized identities (van Zomeren et al., 2008), it is also possible that global

identification would be a weak predictor of action. Given these competing literatures and the

absence of any empirical tests of global identification in the context of collective action, we

consider the role of global identification in understanding Kony2012 action.


7

Another identity alternative is to create an anti-Kony opinion-based group. While

traditional forms of collective action are often underpinned by the involvement of activist or

social movement organizations (SMOs; see Stürmer & Simon, 2004), one of the notable

features of (so-called) modern forms of protest is the absence of formal SMOs. Indeed, as

Bennett and Segerberg (2012) point out, some of the waves of modern protest explicitly

eschew the involvement of formal advocacy organizations because they are seen to be part of

the problem (e.g., Spain’s ‘los indignados’ M15). Opinion-based groups are defined only by

shared values or opinions (Bliuc, McGarty, Reynolds, & Muntele, 2007; McGarty, Bliuc,

Thomas, & Bongiorno, 2009). In this case the relevant opinion is that the campaign to oppose

Kony must be supported. Theoretically, the difference between merely holding an opinion or

attitude and identifying with an opinion-based group is that, in the latter case, people with the

same opinion are perceived to be on the same team; that is, people experience the opinion as a

meaningful basis for collective self-definition (social identification; Turner et al., 1987).

Recently, McGarty et al. (2014) have shown how the explosive growth of mass

oppositional movements in North Africa can be understood as rapidly growing opinion-based

groups organized around the expression of dissent through social (YouTube) and traditional

media (satellite television). Elsewhere, Thomas, Mavor and McGarty (2012) showed that

opinion-based groups predict support for humanitarian collective action on global poverty

amongst members of advantaged groups. Opinion-based groups are not inherently politicized

but Bliuc, McGarty, Hartley and Muntele Hendres (2012) have shown how such groups can

become aligned with nationalist rhetoric and, in doing so, politicize. As such, opinion-based

groups can form potent bases for challenges to authority based on solidarity between

advantaged group members and disadvantaged group members (McGarty et al., 2009; cf.

Subašić, Reynolds & Turner, 2008) or ally-activism (Russell, 2011). Arguably, the Kony2012

video strongly emphasizes the opinion-based nature of the collective identity. The opening


8

and closing sequences [respectively] state that “There is nothing so powerful as an idea whose

time [has come]/ [is now]” (emphasis added). The non-partisan nature of the movement was

represented by the slogan: “One thing we can all agree on”. The current research thus

considers the role of these two identity alternatives (global and opinion-based social

identities) in capturing the global phenomenon of Kony2012.

Contemporary research on collective action also emphasizes the importance of

perceptions of injustice (in particular feelings of anger) and beliefs about efficacy of a cause

(van Zomeren, et al., 2008; see also Alberici & Milesi, 2013; Becker, 2012). The Kony2012

material also explicitly dealt with these factors. Enslaving children is a clear violation of

moral standards, and Kony was presented at the top of a global list of criminals indicted by

the International Criminal Court. The content of the film represented his continued freedom

and activities as a clear case of global injustice. The film also emphasized efficacy by

celebrating the past success of Invisible Children’s campaigns and suggested specific

mechanisms to ensure future success (Waldorf, 2013). In short, it seems that content of the

video drew upon the three factors – social identity, feelings of injustice, and efficacy – that,

according to recent models of collective action are expected to impact on commitment to

action (see van Zomeren et al., 2008, for a meta-analysis; Becker, 2012).

Social Identity: Emergent or Transformed?

Of the three pathways that it specifies, the social identity model of collective action

(SIMCA; van Zomeren et al., 2008) places social identity at the heart of collective

mobilization. According to SIMCA, social identification has a direct relationship with

collective action; but it also has an indirect relationship with action because it stimulates

collective appraisals of injustice and group efficacy. From the perspective of SIMCA, the

Kony2012 video would have triggered a contextually meaningful social identity that

promoted the group-based experience of injustice, efficacy, and consequently, action. In this


9

way, the identity provides the basis for the experience of emotional (injustice) and

instrumental (efficacy) approaches to coping (van Zomeren et al., 2004).

We are also interested here in an alternative set of relationships anticipated by the

encapsulated model of social identity in collective action (EMSICA: Thomas et al., 2009;

2012). EMSICA also places social identity at the heart of collective action but suggests that

social identification may emerge from the appraisal of injustice and group efficacy beliefs

(see also van Zomeren et al., 2008, who anticipate this causal direction). EMSICA therefore

anticipates indirect effects of feelings of injustice and efficacy on action through social

identification. From the perspective of EMSICA, viewing the Kony2012 video would have

triggered feelings of injustice (anger or outrage) and a belief in the efficacy of co-ordinated

action; these then formed the basis for the emergence of a contextually meaningful social

identity. The current research goes beyond a consideration of the nature of the social

identities implicated here (global and opinion-based social identities) to consider the function

of those identities in Kony2012 (transforming or emergent).

The Current Research

This analysis aims to develop a social psychological analysis of global internet

phenomena like Kony2012 by answering three key questions. First, are there differences

between supporters and non-supporters in terms of the features identified by the social

psychological literature on collective action? Scholars have been at pains to differentiate these

modern forms of protest from the more traditional forms of action that have typically been

considered in the collective action literature (see Bennett & Segerberg, 2011, 2012; Castells,

2012, for discussions). If this is the case, then we would not necessarily expect supporters to

be characterised by those attributes that predict traditional socio-political action (e.g. identity,

injustice, efficacy; van Zomeren et al., 2008). On the other hand, if Kony2012 action is

describable in the terms of traditional socio-political action, we would expect supporters to


10

feel angrier, have stronger collective efficacy beliefs and be more willing to take action, than

non-supporters. Given the widespread publicity questioning the legitimacy of the campaign,

we would also expect that those who remained supportive were able to overcome their doubts

while the non-supporters were not (e.g. Haslam & Reicher, 2012).

Moreover, if there are indeed differences between supporters and non-supporters then

that would suggest that it is worth studying the supporters more closely. Therefore, a second

goal is to consider the nature (global or opinion-based) and function (transforming or

emergent) of social identity for supporters of the Kony2012 campaign. Given the theoretical

centrality of social identity in coordinated social behaviour (Turner, et al., 1987) we ask

whether the response to Kony2012 is better described as the mobilization of global social

identification or the formation of an emergent opinion-based group. In other words, is this a

phenomenon most associated with appeals to a (pre-existing) global identity that becomes

subjectively transformed as people come in to contact with the Kony2012 movement; or is the

phenomenon associated with the emergence of a new (opinion-based) identity that comes to

capture initial reactions of injustice and efficacy? Our analysis seeks to determine which set

of causal relations and which identities better describe Kony2012 action.

Third, since much of the momentum of the Kony2012 campaign was generated online

through the posting and sharing of the YouTube clip on social media, we ask: do the same

relationships hold for Kony2012 action generally as compared with actions taken specifically

through social media (Facebook and Twitter)? While some research has considered the

distinction between online and offline actions more generally, results are mixed. For example,

Soon and Kluver (2014) found that political blogging (commonly viewed as a “narcissistic

activity”, p. 500) promoted the emergence of collective identity which drove collective action

both online and offline. However, other research (exploring online socio-political action more

broadly) has suggested that mobilization attempts are specific to media: face-to-face


11

mobilization attempts promote offline actions, while online mobilization attempts promote

online actions (Vissers, Hooghe, Stolle, & Maheo, 2012; see also Brunsting & Postmes, 2002;

Postmes & Brunsting, 2002). Given the centrality of social media in the success generated by

the Kony2012 campaign (and, indeed, new repertoires more generally; Bennett & Segerberg,

2011, 2012), the current research compares the predictors of engagement in Kony2012 social

media action with Kony2012 action overall. We also include a behavioural measure (i.e.

contacting a political decision maker) to compare findings of the self-reported Kony2012

action/s with an observable, traditional socio-political action.

Method

Shortly after the release of the Kony2012 video we collected three separate samples of

survey responses from people present in the public areas of the campuses of two Australian

universities. Wave 1 was two weeks before the day of action; Wave 2 was on the day of

action (April 20); Wave 3 was two weeks after the day of action. Wave 1 began a month after

the initial surge of interest and two weeks after negative publicity in relation to the conduct of

the director of the video. Demographic details (age and gender) for participants across the

three waves are displayed in Table 1.

Measures

Participants were approached by the researchers and asked if they had heard of the

Kony2012 campaign. If they had, they were given some written information about the study

and asked if they would like to participate in exchange for a small bar of chocolate. Of the

people approached, only a minority had not heard of the campaign (~10%); approximately

70% of people approached were prepared to participate. The questionnaire, titled “Attitudes

towards Kony2012”, involved a constant set of items with additional items in Wave 3 to

assess involvement in the day of action. The full set of measures is available from the authors

but the measures we focus on here are presented below.1 Except where indicated all items


12

were measured on 7 point scales (typically anchored 1 “Strongly Disagree” to 7 “Strongly

Agree”).

The key measures were:

Campaign legitimacy. In order to account for the role of the breakdown of legitimacy

of the campaign in movement (de)mobilization, two items measured the perceived campaign

legitimacy: “How legitimate do you believe the focus of the Kony campaign to be?”, “How

legitimate do you believe the Kony campaign to be?”, with Cronbach’s α at all waves > .90.

Anti-Kony attitudes. Attitudes were measured with two items: “Joseph Kony should

be arrested so that he can be charged with war crimes” and “Joseph Kony should be

prosecuted with the full force of international law”, with Cronbach’s α at all waves > .84.

Anger toward Kony. Affective reactions to injustice tend to be better predictors than

cognitive reactions (van Zomeren et al., 2008) so we measured injustice with three items

assessing the emotional reaction of anger: “Thinking about the situation in Uganda and

Joseph Kony’s continued freedom, I feel [Outraged/ Angry/ Fired-up]” with Cronbach’s α >

.84.

Group efficacy. This was measured with two uncorrelated items of which only the

first item (“Together Kony2012 supporters can bring Joseph Kony to justice”) was strongly

related to other theoretically relevant constructs and was retained.2

Anti-Kony opinion-based social identification. Participants were first asked to

respond to an item which assessed nominal opinion-based group membership. Participants

responded “yes” or “no” to the question: “Do you think of yourself as someone who supports

the Kony2012 campaign?”

This categorical item was followed with five questions (one item from each of the five

factors in Leach et al.’s, 2008, scale) which assessed the degree of social identification with

the anti-Kony opinion-based group. The items were: “I feel a bond with other supporters of


13

the Kony2012 campaign”, “I often think about the fact that I am a supporter of the Kony2012

campaign”, “I have a lot in common with other supporters of the Kony2012 campaign”, and

“I am glad to be a supporter of the Kony2012 campaign”. Although these items are intended

to comprise a multidimensional measure the five items held together as a single scale, α > .81

across waves, that we have retained for comparability with the (unidimensional) global

identification measure.

Global social identification. The three item scale developed by Buchan et al. (2011)

was used to measure global social identification, e.g, “How strongly do you feel attachment to

the world as a whole”, α > .83.

Kony 2012 self-reported action. Contemporary self-reported Kony2012 action was

measured on seven items in all three waves. Participants read that: “The Kony campaign calls

for supporters to send messages to prominent people identified on their web site and to join an

international day of action on April 20”. They then indicated whether they had engaged in a

range of Kony2012 actions by responding (yes/no) to the following items: “I have already

posted about Kony on Facebook or Twitter”, “I have already made a donation to the

campaign”, “I have already bought the Kony2012 kit”, I have already talked to friends and

family about the Kony2012 campaign”, “I have already contacted a government official

outlining my support”, “I have already sent a message to one of the people listed on the

Kony2012 website outlining my support”, “I have already made a pledge” (signing a pledge

was intended to indicate commitment to Invisible Children). In Wave 3 we also asked “Did

you participate in the day of action on the 20th April?” Note that we also asked about

intentions to perform each of these behaviours (see Ajzen, 1991); the results of these analyses

are reported in supplementary analyses.

Table 1 shows that, with the exception of two items assessing posting about Kony on

Facebook or Twitter (‘posting’), and talking to friends and family (‘talking’), the reported


14

incidence of the other self-reported actions was extremely low (with only a handful of people

completing the other actions across wave). Accordingly, we created a composite variable

(‘other actions’) that was scored one if any one or more of the five low-incidence actions

(denoted with * in Table 1) was true. We retained the ‘posting’ and ‘talking’ variables so that

we ultimately had three indicators of self-reported Kony action (‘posting’, ‘talking’ and the

‘other actions’ variable); these formed a latent variable of overall Kony self-reported action.

Given that posting on social media (Facebook, Twitter) was the prototypical ‘new repertoire’

and Kony2012 action, and our interest in comparing the different forms of action, we used the

‘posting’ variable as an indicator of Kony social media action specifically.

Kony observed action. As a behavioural measure of a traditional socio-political action

relevant to the Kony2012 campaign, participants were presented with copies of unsigned form

letters to the Australian Foreign Minister supporting or opposing efforts to bring Kony to

justice. Participants could sign one or other letter if they chose. There were 7 people who

signed only the opposition letter and 17 participants who signed both letters. These 17

participants were coded as not signing the supporting letter as we assumed they had

misunderstood the measure but they were retained in all other analyses.

Results

Preliminary Analyses and Statement of Strategy

5 respondents did not complete the survey, had large amounts of missing data and

were therefore removed from the analysis leaving N = 299. Table 1 shows the descriptive

statistics for the key variables at the three waves and correlations are shown in Table 2. Note

that we display the descriptive statistics for the raw self-reported action variables rather than

the latent variable. Apart from a significant reduction in efficacy associated with Wave 3, F

(2, 287) = 3.42, ηp 2 = .02, p = .03, there were no differences in means or frequencies between

waves (all ps .10-.91). The lack of differences across the three waves likely reflects the fact


15

that doubts about the legitimacy of the campaign had already surfaced by the time we took the

initial (Wave 1) measurement 2 weeks before the Cover the Night action: that is, the

campaign had already peaked and plateaued by the time we conducted our Wave 1

measurement. The chart of Google trends (number of people searching Kony) displayed by

Curtis and McCarthy (2012) shows a spike of activity between the 4th and 11th of March but a

flat line of activity thereafter (including at the time of Wave 1 measurement on the 6th April).

We have nevertheless taken account of wave in the analyses below.

Note that our data are comprised of both supporters and non-supporters of the

movement. We consider the differences between supporters and nonsupporters, and the role

of global identification using the entire data set, however, it does not make sense to include

measures of degree of identification (in our tests of SIMCA and EMSICA) for those who do

not identify as supporters. Accordingly, the tests of the function of identification involve only

self-defined supporters of the movement (n = 163).

Main Analyses

Are there differences between Kony2012 supporters and non-supporters? Of the

299 participants, 163 described themselves as supporters of the anti-Kony movement by

selecting ‘yes’ on the nominal group membership item and 136 either indicated that they did

not or failed to respond (n = 28; these people were retained in the analysis as ‘non-

supporters’). To explore our first question, we explored the differences between supporters

and non-supporters using 2 (group membership) X 3 (wave) ANOVA in terms of the

dependent variables campaign legitimacy, anger, efficacy, and global social identification,

and the action variables. There were no significant effects involving wave (all main and

interaction effects were p > .09) except for one marginal group membership by wave

interaction effect on group efficacy, F (1, 275) = 2.91, p = .056, ηp2 = .02. Supporters had

stronger efficacy beliefs as evidenced in the main effect, Ms = 4.39 vs. 2.71, F (1, 275) =


16

73.53, p < .001, ηp2 = .03, but this difference dropped after the day of action (Wave 3), Ms =

3.90 vs. 2.74.

Supporters expressed fewer doubts than did non-supporters about the legitimacy of the

campaign, Ms = 4.43 vs. 2.95, F (1, 275) = 75.05, p < .001, ηp2 = .21, held stronger anti-Kony

attitudes, Ms = 6.41 vs. 5.63, F (1, 284) = 27.90, p < .001, ηp2 = .09, and felt greater anger

toward Kony, Ms = 5.07 vs. 4.11, F (1, 274) = 27.31, p < .001, ηp2 = .09. There were no

differences in global identification, Ms = 4.52 vs. 4.32, F (1, 288) = 1.37, p = .24.

Self-reported action (‘posting’, ‘talking’ and ‘other action’) and observed action were

measured on binary variables so we used logistic regression with the same group membership

(supporter/non-supporter) and wave variables including a centered interaction term to test for

effects. Consistent with the findings above, supporters were more likely to have posted on

social media, 33.1% vs. 14.5%, W (1) = 12.68, p < .001, Exp B = 2.92, talked to friends and

family, 67.7%% vs. 40.9%, W (1) = 21.36, p < .001, Exp B = 3.15, undertaken one of the

other Kony actions, 18.0% vs. 7.7%, W (1) = 6.06, p = .01, Exp B = .38, and completed the

behavioural measure (letter signing), 39.2% vs. 18.2%, W (1) = 14.27, p < .001, Exp B =

2.89; there were no effects of wave and the interaction was not significant for any of the

action variables. We conclude that there are indeed differences between Kony supporters and

non-supporters on the key variables identified in the collective action literature. The

supplementary analyses report the results of a discriminant function analysis showing that

these measures also differentiate supporters from non-supporters3.

What is the nature of social identification in Kony2012? Table 2 helps us to

differentiate between attitudes, opinion-based social identification and global identification.

As can be seen in the table, there is a significant, moderate sized correlation (Cohen, 1988)

between global identification and opinion-based identification (for supporters) but there are

weak correlations between the attitude measure and the two identification measures. This


17

pattern of results suggests that these constructs can be clearly distinguished: holding an anti-

Kony attitude was not the same as identifying with an anti-Kony opinion-based group, and the

two forms of identification had a divergent predictive pattern suggesting that they are indeed

unique constructs. Table 2 also shows that global identification does not predict self-reported

or observed action and is a weak predictor of anger and efficacy. These weak correlations

refute the utility of global identification in explaining Kony2012. We report a further test of

the role of global identification in supplementary analyses however we focus on opinion-

based social identification in our tests of the function of social identification.

What is the function of social identification in Kony2012? To address the third

research question, we conducted a series of multigroup structural equation models (SEM)

using Amos 22. Wave was the grouping variable and we modeled effects on three action

outcomes: self-reported Kony action generally (latent factor), social media action specifically

(posting on Facebook and Twitter), and the observed (behavioural) measure of Kony2012

traditional action. We included only self-defined supporters of the movement (n = 163).

Employing a multigroup SEM allows us to test for the possibility that there were differences

in correlations across the three waves; that is, beyond mean level differences, it allows us to

consider whether the predictors are the same across the three measurement points.

To assess model fit we report several widely accepted goodness-of-fit indices.

Commonly reported indices include Bentler’s (1990) comparative fit index (CFI), a

nonnormed index which compares the given model with a null model. Good fit is shown for

the CFI when the values are greater than .95 (Hu & Bentler, 1999). Second, the root mean

square error of approximation (RMSEA) corrects for model complexity and shows adequate

fit when it is .08 or less (Hu & Bentler, 1999). Finally, we also report the Akaike Information

Criterion (AIC), which is useful when comparing models that are not nested; a smaller value

represents a better fitting, parsimonious model. Table 3 shows the fit statistics for the six


18

models. For all six models we found acceptable fit for models where weights, intercepts and

residuals were constrained to be equal across the three waves, suggesting that there were no

differences in the relationships across wave.

Kony2012 action. We first tested SIMCA, including opinion-based social

identification, anger and group efficacy as direct predictors of action, with anger and efficacy

mediating the indirect effect of identification on action (see Figure 1). Table 3 shows that the

fit statistics were excellent and Figure 1 shows that all the paths were significant except for

the path between efficacy-action. Bootstrapping with 10,000 resamples revealed that the

standardized indirect effect (IE) of identity on action through anger was significant (IE =

0.04, SE = 0.01, 95% CI = 0.01, 0.07), whereas the parallel effect for efficacy (IE = -0.03, SE

= 0.02, 95% CI = -0.08, 0.01) was not significant. We next tested EMSICA, including

opinion-based social identification as a direct predictor of action, with social identification

mediating the indirect effect of group efficacy and anger on action. Table 3 shows that the fit

statistics were good; see Figure 2. Tests of indirect effects showed that the indirect effect of

anger on action through identity was significant (IE = 0.04, SE = 0.02, 95% CI = 0.02, 0.09),

as was the effect of efficacy on action through identity (IE = 0.07, SE = 0.02, 95% CI = 0.04,

0.12). Comparing the AIC of the two self-reported models suggests that the SIMCA model

may be more parsimonious (Table 3) but given the small magnitude of the difference (less

than the rule of thumb nominated by Burnham and Anderson, 2002, of 2.0) we conclude that

both versions of EMSICA and SIMCA fit the data. That aside, anti-Kony social identification

is the strongest predictor of self-reported action.

Kony2012 social media action. We tested the same models using self-reported posting

in social media (Facebook and Twitter) as the dependent variable (see Table 3). The model

fitted the data well although Figure 3 (values to the left of the backslash) shows that the

efficacy-action path was again non-significant and the anger-action path was marginal. The


19

indirect effect through anger was significant (IE = 0.21, SE = 0.08, 95% CI = 0.08, 0.40), but

not for efficacy (IE = -.0.19, SE = 0.12, 95% CI = -0.04, 0.45). EMSICA also fitted the data

well and Figure 4 (values to the left of the backslash) indicates that the paths were all

significant. The indirect effect of anger through identity was not significant (IE = 0.10, SE =

0.07, 95% CI = -0.009, 0.27), but the effect of efficacy through identity was (IE = 0.22, SE =

0.09, 95% CI = 0.07, 0.42). Comparing the AIC suggests that SIMCA was again slightly more

parsimonious.

Observed action. Regarding the measure of observed action (letter signing), the

SIMCA model showed adequate fit, although the chi-square was significant and the RMSEA

was borderline (Table 3). Figure 3 (values to the right of the backslash) shows that the direct

path predicted by SIMCA from efficacy to action was nonsignificant (p = .36) and the path

from anger to action was marginal (p = .06). The indirect effect of identity on action was not

significant for anger (IE = 0.04, SE = 0.08, 95% CI = -0.08, 0.22) or efficacy (IE = 0.02, SE =

0.12, 95% CI = -0.21, 0.28). On the other hand, Table 3 shows that EMSICA demonstrated

good fit and Figure 4 (values to the right of the backslash) shows that all the paths were

significant. Tests of the indirect effects showed that the indirect effect of anger on action

through identity (IE = 0.15, SE = 0.06, 95% CI = 0.04, 0.30) and the effect of efficacy on

action through identity (IE = 0.16, SE = 0.08, 95% CI = 0.02, 0.34) were both significant.

This is consistent with the effects of efficacy and anger being encapsulated by social

identification. The AIC of the two models is the same.

Regarding the question of whether the same relationships applied to Kony2012 action

generally as compared to engagement in social media action, it is striking that a very similar

pattern of results prevailed across the two self-reported forms of action (Kony action and

social media action). We found slightly better fit for SIMCA in both models (as demonstrated

by the AIC; Table 3) but the SIMCA tests also contained a non-significant path from efficacy


20

to action and, for social media action (Figure 3, values to the left), a marginal path from anger

to action. These non-significant paths are more consistent with encapsulation. Indeed, for the

observed action measure, the tests of SIMCA also contained one marginal and one non-

significant path and this also suggests better support for EMSICA. Nevertheless, we can

conclude that the same factors that were implicated in self-reported Kony2012 action

generally applied also to an observable traditional socio-political action (that is, contacting a

political decision maker) and to those actions associated with the ‘new repertoires of protest’

(that is, posting on social media; Polletta, 2011).

Discussion

There is widespread consensus within the social psychological literature that social

identity is central to understanding traditional collective mobilization but some scholars have

suggested that social identity is no longer relevant to modern forms of mobilization (Earl &

Kimport, 2011). The current research sought to develop an analysis of a global Internet

phenomenon in psychological terms.

Results showed that supporters of Kony2012 were markedly different from non-

supporters on the dimensions identified by the traditional collective action literature: they

held stronger anti-Kony attitudes, greater anger towards Kony, believed in the effectiveness of

action and took more (self-reported and observed) action. In view of the widespread criticism

of the campaign after its initial surge of popularity, it is noteworthy that supporters of the

campaign were less likely to express doubts about the legitimacy of the campaign. It is likely

that some of the non-supporters would have classed themselves as supporters if they had not

developed doubts based on that criticism. It is unfortunate that, due to the viral nature of the

campaign and the need for appropriate ethical review, we were not able to commence data

collection until after criticisms had emerged and are therefore unable to empirically

disentangle these processes.


21

Moreover, although previous work identifies global identity as the basis for global

cooperation (Buchan et al., 2011) and global human rights (McFarland et al., 2012, 2013), the

current results suggest that the psychological process underpinning support for Kony2012 is

better understood as commitment to an emergent opinion-based identity (Bliuc et al., 2007).

Consistent with other research which advances a privileged role for identity in collective

action, results pointed to both the transforming and emergent role of identity. We found

support for the idea that the anti-Kony identity acted to promote the expression of anger and

(more tentatively) efficacy, both directly and indirectly promoting action (as per SIMCA; van

Zomeren et al., 2008). We also found support for the encapsulated model of social identity in

collective action (EMSICA) whereby participants first experienced feelings of injustice,

believed in the effectiveness of coordinated action – and these formed the basis for the

emergence of a contextually meaningful identity based on those reactions (Thomas et al.,

2009, 2012).

Although our original intention was to investigate quantitative and qualitative shifts in

the movement over time (i.e. across waves; see Livingstone, 2013), the current results

demonstrated a great deal of stability in responses for this period. Indeed, although we

utilized a convenience sample, we found the same pattern of effects across three different

samples, creating more confidence in our findings. Nevertheless, the design tests the

relationships at specific time points and does not establish a causal direction. Future research

might seek to complement the correlational findings here with experimental evidence that can

test the two pathways in more detail. More generally, replication is important (Yong, 2012)

and it is heartening that we found support for a similar pattern of factors to those documented

in engagement in social movements (e.g., Mazzoni, van Zomeren & Cicognani, in press; but

see Stürmer & Simon, 2009, on anger), traditional forms of collective action (e.g. van


22

Zomeren et al., 2008) and solidarity-based action (van Zomeren, Postmes, Spears & Bettache,

2011; Thomas et al., 2012; c.f. Subašić et al., 2008).

Finally, we found that engagement in social media action (sharing Kony2012 material

on Facebook or Twitter), engagement in Kony2012 action generally and an observed

traditional socio-political action (letter signing) were predicted by the same factors. In short,

the results suggest that social media action was associated with a meaningful and valid

expression of collective selfhood (social identity); this seems unlikely to have been the case if

engagement with the cause is transient and of little enduring meaning to the participants (as

was a core criticism of the Kony2012 campaign: Mengestu, 2012). Indeed, the results are

consistent with other emerging research that indicates that social media use facilitates political

and civic engagement, both online and offline (de Zúñiga, Jung & Valenzuela, 2012; Soon &

Kluver, 2014), and particularly amongst those who might otherwise disengage (Chan & Guo,

2013; Shah, Cho, Eveland & Kwak, 2005). In what follows we consider the implications of

these results for developing an analysis of the nature and function of identity in new models

and methods of social change (see Castells, 2012).

The Nature and Function of Identity in New Social Media

The results support the idea that movements such as Kony2012 find force by drawing

on an orientation about the way the world should be, and creating a sense of solidarity with

others who share that world view (McGarty et al., 2009; McGarty et al., 2014). As such it is

consistent with previous evidence implicating the utility of opinion-based groups in

understanding ‘ally activism’ (Russell, 2011; or solidarity-based action, see also Subašić et

al., 2008; van Zomeren et al., 2011) and humanitarian action amongst advantaged group

members (Thomas et al., 2012). One of the strengths of opinion-based groups is that they

transcend social categorical group boundaries (e.g. national, gender or ethnic boundaries;

McGarty et al., 2009) such that members of both advantaged and disadvantaged groups can


23

both identify with an opinion about ‘how the world should be’ (Smith et al., in press). In this

way, opinion-based groups seem well placed to capture emergent (new, novel) social

movements such as Kony2012 because they can capture perceived discrepancies between

how the world ‘is’ and how it ‘should be’ (McGarty et al., 2014; Smith, Thomas, & McGarty,

in press).

On the other hand, we found little support for the role of global identity in these data.

As such, the current findings stand in contrast to those of McFarland et al. (2012, 2013) who

show that a superordinate identification with humanity promotes interest in, and engagement

with, global human rights issues. One possibility is that global social identification is more

relevant to cooperation and public goods dilemmas of the form explored by Buchan et al.

(2011). Another difficulty, though, is that who belongs to – and is prototypical of – the global

psychological community may be hotly contested (Reese et al., 2012; also Greenaway et al.,

2011; Morton & Postmes, 2011). Thus, superordinate human and global identities may be

inherently more difficult to associate with specific norms for action (Thomas et al., 2009).

Nevertheless, future research might further consider the role of other identities

(including global identity) as people interact with movements, challenge powerful outgroups

(see Drury & Reicher, 2000), and come to adopt new norms and values that engender

productive forms of global citizenship. Indeed, it is possible that other campaigns may more

successfully draw on a pre-existing identity (global or otherwise) and rhetorically link

injustice and efficacy as the ‘work’ of group members (see Bliuc et al., 2012; Reicher &

Hopkins, 2001; Reicher, Cassidy, Wolpert, Hopkins, & Levine, 2006). New emotions and

beliefs may also become more relevant as the movement gains momentum or suffers set-

backs (e.g., Tausch & Becker, 2013). And, given the important role ascribed to morality in

recent developments in the collective action literature (e.g. van Zomeren et al., 2011, 2012;


24

van Zomeren, 2013), future research should consider the role of (absolute) moral positions in

precipitating and sustaining involvement in these forms of movements.

Although we have focused our discussion here on the effects of modern social media,

in many ways attitudes, anger, efficacy and identity are also temporal outcomes of what is a

more distal, communicative, process. That is, much of what makes social media such a

powerful tool for social change is their potential to enable rapid communication across time

and space (Castells, 2012; McGarty et al., 2014, for a case study), and the fact that they

provide social spaces for people to share, discuss and contest their positions (e.g., Soon &

Kluver, 2014; see Thomas, Smith, McGarty & Postmes, 2010 more generally). Indeed,

Bennett and Segerberg (2011) describe modern forms of social mobilization as connective

(not collective) action. Social psychology is well placed to contribute an analysis of social

influence in modern forms of mobilization (following Spears et al., 2002) and, in this vein,

future research should consider aspects of the online environment in which potential

participants encounter and experience mobilizing content (see also Thomas & Louis, 2014,

for an account of collective action as social influence).

New Forms of Mass Mobilization and Social Change

Kony2012 was criticized for its failure to achieve its stated policy objective and

generate significant offline action, giving rise to a rather grim view of modern social

movements (e.g. Gladwell, 2010, for social media more broadly). Consistent with this view,

the average levels of psychological commitment to the anti-Kony group considered here were

not high. That is, although participants held strongly anti-Kony attitudes, the campaign itself

was not strongly self-defining (the average level of identification with the group was weak),

and it may be that their action reflected this. Moreover, the data point to a key limitation of

the campaign. Specifically, key actions requested by the campaign organizers required a very

high commitment from new supporters of the cause. Supporters were asked to purchase a


25

$US30 kit from an online store and to put up posters on the night of April 20 “after sundown”

(creating what Klandermans, 1997, would term a ‘barrier to participation’ for many young

people) without identifying specific locations. We suspect that if the organizers of the

Egyptian protests of 2011 had expected protesters to pay to participate, and had neglected to

suggest that protesters assemble in a specific place (Tahrir Square), then they would have

failed to generate action too.

However, rather than denigrating modern forms of mobilization we believe our data

also supports an alternative, more positive, view. Rather than pre-judge actions as effective or

ineffective (c.f. Hornsey et al., 2006), a perspective grounded in social interactionism would

ask how participants subjectively understand and evaluate their actions (e.g. Turner & Oakes,

1986). Our data showed that similar processes underpinned engagement in a range of self-

reported actions, observed action, and an action characteristic of new repertoires of protest

(posting on social media). The Kony2012 video is juxtaposed with images of people taking

action online, as the narrator implores “The better world we want is coming, it’s just waiting

for us to stop at nothing” (Invisible Children, 2012). There is therefore good reason to think

that participants believed that, by sharing the YouTube video or ‘liking’ it on Facebook, they

were directly supporting a campaign to capture a war criminal (Waldorf, 2013). Kony2012

exposed people to a powerful delegitimizing narrative and presented them with achievable

action, allowing them to overcome uncertainty and act to embody imagined cognitive

alternatives, all vital components of a psychology of resistance (Haslam & Reicher, 2012).

Judging the effectiveness of modern movements based on the yardstick of traditional methods

(rallies, petitions) will limit the potential of the social and behavioural sciences to develop

explanations of these phenomena.

Writing on the topic of the Occupy movement, Castells (2012, p.193) notes that

Occupy did not achieve widespread policy change but that it “shifted public discourse”. One


26

of the protesters put it differently, writing on the Occupy Facebook page: “The process is the

message” (see Castells, 2012, p. 203, italics added). Given that policy change has been linked

to public opinion (e. g., Burstein, 2003; see also Louis, 2009) a more positive view of the

achievements of modern forms of social mobilization seems well-founded. Indeed, the current

analysis suggests a nuanced view of the effect and impact of social media on social change.

The use of social media entails real communication between real people: it cannot be ignored

as part of some other “virtual” world (Jurgenson, 2012). The fact that around 30% of

participants were prepared to sign a letter to a government minister points to the (untapped)

power of the campaign to create impressive levels of conventional political action alongside

online interaction.


27

Footnote

1 Other measures focused on the perceived locus of responsibility, norms of opinion and

action and degree of political trust within the national (Australian) and global community. For

those participants that reported having taken action, we also asked about how that made them

feel (hopeful, proud).

2 The second efficacy item read: “Efforts to bring Kony to justice are a waste of everyone's

effort” (reverse scored). Compared to the magnitude of the correlations of the first item (see

Table 2; rs = .31-.69), this item was only weakly correlated with action, anti-Kony attitudes,

anger and anti-Kony identification (rs = .14-.17).

3 For completeness we also ran the wave X group analyses excluding the 28 participants who

did not answer the question about opinion-based group membership. The pattern of main

effects for group membership was unchanged but there were some additional wave X group

interactions of small size that were significant in the range p = .03 to .08. Given the focus on

opinion-based membership in the next section we have not interpreted these effects.


28

References

Ajzen, I. (1991). The theory of planned behaviour. Organizational behaviour and human

decision process, 50, 179-211.

Alberici, A. I., & Milesi, P. (2013). The influence of the Internet on the psychosocial

predictors of collective action. Journal of Community and Applied Social Psychology,

23, 373-388.

Bailyn, E. (2012). The difference between slacktivism and activism: How ‘Kony2012’ is

narrowing the gap. Huffington Post. Retrieved 15 August 2013 from

http://www.huffingtonpost.com/evan-bailyn/kony-2012-activism_b_1361791.html

Bandura, A. (2000). Exercise of human agency through collective efficacy. Current

Directions in Psychological Science, 9, 75-78.

Becker, J. C. (2012). Virtual special issue on theory and research on collective action in the

European Journal of Social Psychology. European Journal of Social Psychology, 42,

19-23.

Bennett, L., & Segerberg, A. (2011). Digital media and the personalization of collective

action: Social technology and the organization of protests against the global economic

crisis. Information, Communication & Society, 14, 770-799.

Bennett, L., & Segerberg, A. (2012) The logic of connective action. Information,

Communication & Society, 15, 739-768.

Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin,

107, 238-246.

Bliuc, A. M., McGarty, C., Reynolds, K., & Muntele, D. (2007). Opinion-based group

membership as a predictor of commitment to political action. European Journal of

Social Psychology, 37, 19-32.


29

Bliuc, A.-M., McGarty, C., Hartley, L. K., & Muntele Hendres, D. (2012). Manipulating

national identity: The strategic use of rhetoric by supporters and opponents of the

'Cronulla riots' in Australia. Ethnic and Racial Studies, 35, 2174-2194.

Brunsting, S., & Postmes, T. (2002). Social movement participation in the digital age:

Predicting offline and online collective action. Small Group Research, 33, 525-554.

Buchan, N. R., Brewer, M. B., Grimalda, G., Wilson, R. K., Fatas, E., & Foddy, M. (2011).

Global social identity and global cooperation. Psychological Science, 22, 821-828.

Burnham, K. P., & Anderson, D. R. (2002). Model selection and multimodel inference: A

practical information-theoretic approach (2nd Ed). New York: Springer-Verlag./

Burstein, P. (2003). The impact of public opinion on public policy: A review and an agenda.

Political Research Quarterly, 56, 29-40.

Castells, M. (2012). Networks of outrage and hope: Social movements in the internet age.

Polity Press: Cambridge, UK.

Chan, M., & Guo, J. (2013). The role of political efficacy on the relationship between

Facebook use and participatory behaviors: A comparative study of young American

and Chinese adults. Cyberpsychology, behavior, and social networking, 6, 460-463. Cohen, J. (1988). Statistical Power Analysis for the Behavioral Sciences (2nd Ed.). Hillsdale,

NJ: Erlbaum.

Curtis, P., & McCarthy, T. (2012). Kony2012: What happens next? The Guardian. Retrieved

15 August 2013 from http://www.guardian.co.uk/politics/reality-check-with-polly-

curtis/2012/apr/20/kony-2012-what-happens-next.

de Zuniga, HG., Jung N., & Valenzuela, S (2012). Social media use for news and individuals’

social capital, civic engagement and political participation. Journal of Computer-

Mediated Communication, 17. 319–336.

Drury, J., & Reicher, S. (2000). Collective action and psychological change: The emergence


30

of new social identities. British Journal of Social Psychology, 39, 579-604.

Earl, J., & Kimport, K. (2011). Digitally enabled social change: Activism in the internet age.

Massachusetts Institute of Technology: Cambridge, MA.

Gladwell, M. (2010). Small change: Why the revolution will not be tweeted. New Yorker.

Retrieved 15 August 2013 from

http://www.newyorker.com/reporting/2010/10/04/101004fa_fact_gladwell

Greenaway, K. H., Quinn, E. A., & Louis, W. R. (2011). Appealing to common humanity

increases forgiveness but reduces collective action intentions amongst victims of

historical atrocities. European Journal of Social Psychology, 41, 569-573.

Haslam, S. A., & Reicher, S. D. (2012). When prisoners take over the prison: A social

psychology of resistance. Personality and Social Psychology Review, 16, 154-179.

Hornsey, M. J., Blackwood, L., Louis, W. R., Fielding, K., Mavor, K., Morton, T., O’Brien,

A., Paasonen, K., Smith, J., & White, K. M. (2006). Why do people engage in

collective action? Revisiting the role of perceived effectiveness. Journal of Applied

Social Psychology, 36, 1701–1722.

Howard, P. N., Duffy, A., Freelon, D., Hussain, M. M., Mari, W., & Mazaid M. (2011).

Opening closed regimes: What was the role of social media during the Arab Spring?

Project on Information Technology and Political Islam, Working Paper 2011.1.

Retrieved March 21, 2012, from http://pitpi.org/?p=1051.

Hu, L., & Bentler, P.M. (1999). Cutoff criteria for fit indexes in covariance structure analysis:

Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1-55.

Invisible Children (2012). Cover the Night. Retrieved November 16, 2012 from

http://ctn.kony2012.com/

Jurgenson, N. (2012). When atoms meet bits: Social media, the mobile web and augmented

revolution. Future Internet, 4, 83-91.

http://pitpi.org/?p=1051

http://ctn.kony2012.com/


31

Klandermans, B. (1997). The social psychology of protest. Oxford: Blackwell.

Leach, C. W., Snider, N., & Iyer, A. (2002). “Spoiling the consciences of the fortunate”: The

experience of relative advantage and support for social equality. In I. Walker & H. J.

Smith (Eds.), Relative deprivation: Specification, development, and integration (pp.

136-163). New York: Cambridge University Press.

Leach, C. W., van Zomeren, M., Zebel, S., Vliek, M. L. W., Pennekamp, S. F., Doosje, B., &

Spears, R. (2008). Group-level self-definition and self-investment: A hierarchical

(multicomponent) model of in-group identification. Journal of Personality and Social

Psychology, 95, 144-165.

Livingstone, A. (2013). Why the psychology of collective action requires qualitative

transformation as well as quantitative change. Contemporary Social Science: Journal of

the Academy of Social Sciences, 9, 121-134.

Lotan, G., Graeff, E., Ananny, M., Gaffney, D., Pearce, I., & boyd, d. (2011). The Arab

Spring. The Revolutions were tweeted: Information flows during the Tunisian and

Egyptian revolutions. International Journal of Communication, 5, 1375-1405.

Louis, W. R. (2009). Collective action - And then what? Journal of Social Issues, 65, 727-

748.

Mazzoni, D., van Zomeren, M., Cicognani, E. (in press). The motivating role of perceived

right violation and efficacy beliefs in identification with the Italian Water Movement.

Political Psychology. DOI: 10.1111/pops.1210

McFarland, S.G., Brown, D., & Webb, M. (2013). Identification with humanity as a moral

concept and psychological construct. Current Directions in Psychological Science, 22,

194-198.


32

McFarland, S. G., Webb, M., & Brown, D. (2012). All humanity is my ingroup: A measure

and studies of identification with all humanity. Journal of Personality and Social


McGarty, C., Bliuc, A.-M, Thomas, E. F., & Bongiorno, R. (2009). Collective action as the

material expression of opinion-based group membership. Journal of Social Issues, 65,

839-857.

McGarty, C., Thomas, E. F, Lala, G., Smith, L. G. E., & Bliuc, A.-M. (2014). New

technologies, new identities, and the growth of mass opposition in the ‘Arab Spring’.

Political Psychology, 35,725-740. DOI: 10.1111/pops.12060

Mengestu, D. (2012). Not a click away: Joseph Kony in the real world. Warscapes. Retrieved

15 August 2013 from http://www.warscapes.com/reportage/not-click-away-joseph-

kony-real-world

Morton, T., & Postmes, T. (2011). Moral duty or moral defence? The effects of perceiving

shared humanity with the victims of ingroup perpetrated harm. European Journal of

Social Psychology, 41, 127-134.

Nadler, A., & Halabi, S. (2006). Intergroup helping as status relations: Effects of status

stability, identification, and type of help on receptivity to high-status group’s help.

Journal of Personality and Social Psychology, 91, 97-110.

News (2012). Kony campaign falls flat in Sydney. Retrieved October 26, 2012 from

http://www.news.com.au/breaking-news/kony-campaign-falls-flat-in-sydney/story-

e6frfku0-1226334899506.

Pew Research Center (2012). The viral Kony 2012 video. Retrieved 26 November 2014 from

http://www.pewinternet.org/files/old-

media//Files/Reports/2012/The_Viral_Kony_2012_Video.pdf.


33

Polletta, F. (November 21 2011). Maybe you’re better off not holding hands and singing we

shall overcome. Mobilizing Ideas. Retrieved 15 August 2013 from

http://mobilizingideas.wordpress.com/2011/11/21/maybe-youre-better-off-not-holding-

hands-and-singing-we-shall-overcome/

Postmes, T., & Brunsting, S. (2002). Collective action in the age of the internet: Mass

communication and online mobilization. Social Science Computer Review, 20. 290-301.

Reicher, S. D. (1984). The St. Pauls riot: An explanation of the limits of crowd action in

terms of a social identity model. European Journal of Social Psychology, 14, 1–21.

Reicher, S., Cassidy, C., Wolpert, I., Hopkins, N., & Levine, M. (2006). Saving Bulgaria's

Jews: An analysis of social identity and the mobilisation of social solidarity. European

Journal of Social Psychology, 36, 49-72.

Reicher, S.D., Hopkins, N., Levine, M., & Rath, R. (2005). Entrepreneurs of hate and

entrepreneurs of solidarity: Social identity as a basis for mass communication.

International Review of the Red Cross, 87, 621-637.

Reicher, S.D., & Hopkins, N. (2001). Self and Nation. Sage: London.

Reese, G., Berthold, A., & Steffens, M. C. (2012). We are the world—and they are not:

Prototypicality for the world community, legitimacy, and responses to global inequality.

Political Psychology, 33, 683–700.

Russell, G. (2011). Motives of Heterosexual Allies in Collective Action for Equality. Journal

of Social Issues, 67, 376 - 393. DOI: 10.1111/j.1540-4560.2011.01703.x

Shah, D.V., Cho, J., Eveland, W., & Kwak, N. (2005). Information and expression in a digital

age: Modeling internet effect on civic participation. Communication Research, 32, 531-

565.

Simon, B., & Klandermans, B. (2001). Politicized collective identity: A social psychological

analysis. American Psychologist, 56, 319-331.


34

Soon, C., & Kluver, R. (2014). Uniting political bloggers in diversity: Collective identity and

web activism. Journal of Computer Mediated Communication, 19, 500-515.

Smith, L. G. E., Thomas, E. F. & McGarty, C. (in press). Forming the identity-norm nexus: A

theory of the interactive psychological underpinnings of social change. Political

Psychology. DOI: 10.1111/pops.12180.

Spears, R., Lea, M., Corneliussen, R. A., Postmes, T., & Ter Haar, W. (2002). Computer-

mediated communication as a channel for social resistance: The strategic side of

SIDE. Small Group Research, 33, 555-574.

Stürmer, S., & Simon, B. (2004). Collective action: Towards a dual-pathway model.

European Review of Social Psychology, 15, 59-99.

Stürmer, S., & Simon, S. (2009). Pathways to collective protest: Calculation, identification or

emotion? A critical analysis of the role of group-based anger in social movement

participation. Journal of Social Issues, 65, 681-726.

Subašić, E., Reynolds, K.J., & Turner, J.C. (2008). The political solidarity model of social

change: Dynamics of self-categorization in intergroup power relations. Personality

and Social Psychology Review, 12, 330-352.

Tausch, N. & Becker, J. (2013) Emotional reactions to success and failure of collective action

as predictors of future action intentions: A longitudinal investigation in the context of

student protests in Germany. British Journal of Social Psychology, 52, 525-542.

doi/10.1111/j.2044-8309.2012.02109.x

Thomas, E. F., & Louis, W.R. (2014). When will collective action be effective? Violent and

non-violent protests differentially influence perceptions of legitimacy and efficacy

amongst supporters. Personality and Social Psychology Bulletin, 40, 263-276. DOI:

10.1177/0146167213510525


35

Thomas, E. F., Mavor, K. I., & McGarty, C. (2012). Social identities facilitate and

encapsulate action-relevant constructs: A test of the social identity model of collective

action. Group Processes and Intergroup Relations, 15, 75-88.

Thomas, E. F., McGarty, C., & Mavor, K. I. (2009). Aligning identities, emotions and beliefs

to create commitment to sustainable social and political action. Personality and Social

Psychology Review, 13, 194-218.

Thomas, E. F., Smith, L. G. E., McGarty, C., & Postmes, T. (2010) Nice and nasty: The

formation of prosocial and hostile social movements. International Review of Social


Time (2012). Top 10 Everything of 2012. Retrieved 15 August 2013 from

entertainment.time.com/2012/12/04/top-10-arts-lists/slide/kony-2012/

Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D. & Wetherell, M. S. (1987).

Rediscovering the social group: A self-categorization theory. Cambridge: Blackwell.

Turner, J. C., & Oakes, P. J. (1986). The significance of the social identity concept for social

psychology with reference to individualism, interactionism and social influence.

British Journal of Social Psychology, 25, 237–252.

van Zomeren, M., (2013). Four core social-psychological motivations to undertake collective

action. Social and Personality Psychology Compass, 7, 378-388.

van Zomeren, M., Leach, C. W., & Spears, R. (2012). Protesters as "Passionate Economists":

A dynamic dual pathway model of approach coping with collective disadvantage.

Personality and Social Psychology Review, 16, 180–199.

van Zomeren, M., Postmes, T., & Spears, R. (2008). Toward an integrative social identity

model of collective action: A quantitative research synthesis of three socio-

psychological perspectives. Psychological Bulletin, 134, 504 –535.


36

Van Zomeren, M., Postmes, T., Spears, R., & Bettache, K. (2011). Can moral convictions

motivate the advantaged to challenge social inequality? Extending the Social Identity

Model of Collective Action. Group Processes & Intergroup Relations,14, 735–753.

Vissers, S., Hooghe, M., Stolle D. & Maheo, V-A. (2012). The impact of mobilization media

on off-line and online participation: Are mobilization effects medium specific? Social

Science Computer Review, 30, 152-162.

Waldorf, L. (2013). White noise: Hearing the disaster. Journal of Human Rights Practice, 4.

469-474.

Yong, E. (2012). Bad copy: In the wake of high profile controversies, psychologists are

facing up to problems with replication. Nature, 485, 298-300.


37

Table 1. Means (Standard Deviations) of Key Variables by Wave

Wave 1 Wave 2 Wave 3

N 95 102 102

Age 20.67 (4.55) 20.26 (4.18) 21.27 (5.31)

% Female 45.3 60.8 48.0

Attitudes 6.02 (1.38) 6.03 (1.50) 6.17 (0.96)

Anger 4.62 (1.63) 4.78 (1.55) 4.57 (1.52)

Group efficacy 3.92 a (1.84) 3.82 a (1.92) 3.28 b (1.73)

Global identification 4.60 (1.34) 4.51 (1.38) 4.17 (1.44)

Anti-Kony identification 3.95 (1.14) 3.52 (1.11) 3.62 (1.17)

Perceived legitimacy 3.84 (1.58) 3.63 (1.51) 3.90 (1.65)

% Observed action 30.3 33.0 25.3

% Posted about Kony 23.2 25.5 24.5

% Talked to friends, family 52.6 57.8 53.9

% Donated* 6.3 3.9 6.9

% Purchased kit* 6.3 3.9 1.0

% Made a pledge* 10.5 2.9 6.9

% Contacted official* 2.1 2.9 2.9

% Message Kony targets* 3.2 2.0 2.0

% Day of Action -- -- 2.0

Notes. Cell sizes vary with missing data and, in the case of the observed action, those who

signed both letters were removed from the analysis. For anti-Kony social identification values

are for supporters only (n = 163). * Denotes those self-reported variables that were combined

to form the ‘other action’ variable. Means/frequencies with different subscripts differ at p <

.05.


Table 2. Correlations of Key Dependent Variables across Wave (N = 299)

1 2 3 4 5 6 7 8 9 10 11

1.Group membership - .28*** .30*** .46*** … .08*** .45*** .21*** .27*** .15*** .23***

2. Attitudes - .53*** .31*** .10 .11 .31*** .22*** .26 .13* .15*

3. Anger - .44*** .42*** .13* .43*** .44*** .43** .20** .30***

4. Group efficacy - .59*** .19*** .54*** .44*** .51*** .11* .28***

5. Anti-Kony identification - .35*** .49*** .51*** .58*** .23** .26***

6. Global identification - .11 .16** .19 .08 .02

7. Legitimacy - .44*** .51** .17** .29***

8. Self-reported posting - .61*** .39*** .25***

9. Self-reported talk - .26** .20**

10. Self-reported other action - .38***

11. Observed action -

NB: Group membership coded supporter = 1, non-supporter = -1. For anti-Kony social identification values are for supporters only (n = 163).


Table 3. Fit Statistics for the SIMCA (Figures 1 and 3) and EMSICA (Figure 2 and 4) Models

for Kony Action, Social Media Action and Observed Action.

____________

Fit statistics ________________________

Variable CFI RMSEA χ2 (df) p AIC

SIMCA all action .96 .04 75.48 (55) .04 127.48

SIMCA social media 1.00 .00 24.67 (29) .55 50.67

SIMCA observed 1.00 .08 21.20 (29) .85 47.20

EMSICA all action .95 .04 80.93 (56) .02 130.93

ESIMCA social media 1.00 .00 28.35 (30) .55 52.35

ESIMCA observed 1.00 .00 23.80 (30) .78 47.80

_____________


40

Figure 1. Standardized regression coefficients for the structural model of the social

identity model of collective action predicting Kony2012 action. *** denotes that path is

significant at p < .001. ** denotes that path is significant at p < .01. * denotes that path is

significant at p < .05.

Anger

Anti-Kony identification

Group Efficacy

.45***

.71***

.42***

.56***

Anti-Kony action

Posting

Talking

Other action

.60***

.75***


41

Figure 2. Standardized regression coefficients for the structural model of the

encapsulated model of social identity in collective action predicting Kony2012 action. ***

denotes that path is significant at p < .001, ** denotes that path is significant at p < .01.


Anger

Group Efficacy

.60***

.50***

.70***

.43**

.56***

Anti-Kony action

Posting

Talking

Other action

.33***

.45***


42

Figure 3. Standardized regression coefficients for the structural model of the social

identity model of collective action. Figures to the left of the backslash relate to the values for

Kony social media action; figures to the right of the backslash relate to the value for the

measure of observed action (letter signing). *** denotes that path is significant at p < .001. **

denotes that path is significant at p < .01. * denotes that path is significant at p < .05. † denotes

that path is marginally significant at p = .06.

Anger


Group Efficacy

.22*/.34***

.13†/.11.19*

-.06 /.00

.60*** / .60***

.75*** / .75***

Social media action / Observed action


43

Figure 4. Standardized regression coefficients for the structural model of the

encapsulated model of social identity in collective action. Figures to the left of the backslash

relate to the values for Kony social media action; figures to the right of the backslash relate to

the value for the measure of observed action (letter signing). *** denotes that path is significant

at p < .001.


Anger

Group Efficacy

Social media action / Observed action

.60***/.60***

.26***/.40***

.45***/.45*** NS

.33***/.33*** NS

publisher's layout or pagination. thomas, e.f. , mcgarty ... · lala, smith, & bliuc,...

Documents